The AI Daily Brief
The Multiplayer
AI Sprint
for Teams
Agents today are single-player. Everyone runs their own, in their own tools, for their own tasks, while the actual work happens between people. Over four weeks, your team works out which agents belong at the intersection of the work, and builds one.
Meeting notes
Research
Drafting
Reporting
One shared agent
Agents were built for one person, and almost no job is done by one person.
Your agent knows your tasks and your files. It knows nothing about the handoff to the person downstream of you, which is where the work usually goes wrong.
Everything each of you taught your agent about the product, the customers, and the way things get decided lives in a place nobody else on the team can reach.
Four weeks
Inventory
Find out what your team is actually running.
Your team's inventory, its first reveal, and a voted shortlist of shared use cases.
Context
Write down what your AI knows about you, and what it should.
A private context per person and one shared repository of plain files any agent can read.
The overlap
Decide which shared agent to build first, and design it.
One build target with one owner, the decision written down, and the agent's one-page design.
Shared Agent
Ship one shared agent. Then find out if it was worth it.
A running shared agent, versioned shared context, and an honest verdict.
New to the idea? Read the guide: Understand multiplayer AI →
Start here
Make an account, then create a team and send the code to the people you work with, join a team someone already made, or start on your own and add people later. Teams can be any size; four is a sensible place to start. It works far better with a team; bring one or two colleagues.
Every step comes three ways: do it here with our AI, take a prompt to the AI tool your company already allows, or bring your team's own way. Nothing here needs a new tool.
Questions while you run the sprint? The AI operators' help desk is a community where you can ask people who are building alongside you.
Join the Community Help Desk →